New inhibitors of rabbit muscle triose-phosphate isomerase

Article abstract of DOI:10.1016/j.bmcl.2005.03.061

We describe the synthesis and evaluation of three new competitive inhibitors of triose-phosphate isomerase. One of them (phosphoglycoloamidoxime: K_i = 4.5 μM) is among the best reversible inhibitors so far reported for this enzyme.

Full text of DOI:10.1016/j.bmcl.2005.03.061

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2906–2909
New inhibitors of rabbit muscle triose-phosphate isomerase
M. Fonvielle, S. Mariano and M. Therisod*
LCBB, ICMMO, Bat. 420, Universit e´ Paris-Sud, F-91405 Orsay, France
Received 26 January 2005; revised 8 March 2005; accepted 17 March 2005
Available online 19 April 2005
Abstract—We describe the synthesis and evaluation of three new competitive inhibitors of triose-phosphate isomerase. One of them
(
i
phosphoglycoloamidoxime: K = 4.5 lM) is among the best reversible inhibitors so far reported for this enzyme.
ꢀ
2005 Elsevier Ltd. All rights reserved.
1
. Introduction
relatively weak reversible inhibitors (K = 0.2–1.3 mM).
i
5
Phosphoglycolohydroxamate 1a and phosphoglycolic
6
Triose-phosphate isomerase (TIM; EC 5.3.1.1.) cata-
lyzes the reversible interconversion of dihydroxyacetone
phosphate (DHAP) and glyceraldehyde phosphate
acid are the best competitive inhibitors, with K values
i
of 6–14 and 3 lM, respectively, against the rabbit
muscle enzyme. These two compounds, in spite of their
relatively low affinity for the enzyme (K /K around
(
G3P). Only G3P can proceed further down the glyco-
m
i
lytic pathway (Scheme 1). It is one of the most studied
enzymes.
300), have been considered as Ôhigh-energy intermediate
(HEI) analogueÕ and Ôtransition-state analogueÕ, respec-
tively (Fig. 1). All these compounds are active on
7
TIM from different sources. Only one hexapeptidic
inhibitor shows a marked selectivity (K = 3 lM) against
the Trypanosoma enzyme.
i
8
The synthesis and testing of new inhibitors of TIM are
of interest for the main following reasons:
•
Most of the crystal structures so far reported have
been obtained on TIM-inhibitor complexes. For
future new structures to be resolved, other inhibitors
will be welcome.
Scheme 1. Fructose-bisphosphate aldolase (FBP-aldolase) and TIM:
two central enzymes of the glycolysis.
•
Triose-phosphate isomerase is a ubiquitous glycolytic
enzyme. However, it could be considered as a poten-
tial therapeutic target if significant differences in
TIM catalytic mechanism is for a large part understood,
and has even been considered as a model for enzyme effi-
1
ciency knowledge. Tertiary structures of TIM from dif-
2
ferent sources have been resolved (Plasmodium, rabbit,
yeast, chicken, Trypanosoma, Leishmania, Bacillus,
human, etc.). Surprisingly, only a few inhibitors of
3
TIM have been reported: halo-acetone phosphate and
4
glycidol phosphate were used as irreversible inactiva-
tors to label active site residues. 3-Phosphoglycerate,
glycerol phosphate and phosphoenol pyruvate are
Keywords: Triose-phosphate isomerase; Class II fructose-bisphosphate
aldolase; Enzyme inhibitor; Phosphoglycoloamidoxyme; Phospho-
glycolohydrazide; Phosphonate inhibitors.
*
Corresponding author. Tel.: +33 1 69 15 6311; fax: +33 1 69 15
281; e-mail: therisod@icmo.u-psud.fr
Figure 1. Substrate, high-energy intermediate (HEI) and synthesized
products.
7
0
960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.061

M. Fonvielle et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2906–2909
2907
structures are found between human TIM and the
enzyme from different pathogenic species. Selective
inhibition of the enzyme present in microorganisms
Compound 2a was synthesized according to our previ-
1
6
ously described method. Phosphoglycoloamidoxime
1
7
3a was prepared according to Scheme 3, by a simpli-
fied procedure as compared to our previously described
13
preparation.
9
,10
11
such as Plasmodium,
8
Leishmania or Trypano-
whose growth depends strongly or exclu-
,12
soma,
sively on glycolysis is of special interest and has
already been considered.
•
We recently published the synthesis and evaluation of
two new selective inhibitors of class II FBP-aldolase:
phosphoglycolohydrazide 2a and phosphoglycolo-
3. Results and discussion
Enzymatic assays: the formation of DHAP from G3P
was followed spectrophotometrically (340 nm) by mea-
suring the consumption of NADH in a coupled system
employing TIM and a large excess of glycerol-3-phos-
phate dehydrogenase, according to established proce-
dures (the concentration of G3P in the assay was
estimated enzymatically with the same coupled system).
Not surprisingly, the two phosphonates 2b and 3b were
very poor inhibitors of TIM (Table 1). Only 1b showed a
amidoxime 3a (K = 0.34 and 2.3 lM against yeast
i
1
3
aldolase, respectively). These compounds are very
weak inhibitors of the mammalian FBP-aldolase.
Since class II aldolase is present only in microorgan-
isms (among which many pathogenic bacteria and
yeasts), we consider these inhibitors as potential syn-
thetic antibiotics. This is realistic only if other mam-
malian glycolytic enzymes (like TIM) are not
significantly inhibited by these products.
1
9
significant inhibition, with a K value of 145 lM (to be
i
compared with a published value of K = 3 lM for the
i
5
a
We report hereby the synthesis of 2a, 3a, and of their
corresponding phosphonates analogues 2b, 3b, and their
testing as inhibitors of a mammalian TIM. We also pre-
pared and evaluated compound 1b (the phosphonate
analogue of the old known phosphoglycolohydroxa-
phosphate 1a). This confirms that a phosphoester oxy-
gen is required for an efficient binding of inhibitors (sub-
strates) into the active site of TIM. Occasionally, the
same conclusion was deduced from our parallel experi-
ments with FBP-aldolases (Table 1).
5
14
mate 1a, ), which was briefly described only recently.
On another hand, the two phosphates 2a and 3a were
more effective inhibitors of TIM, with K values of 111
As shown in Figure 1, these products are (at least) ana-
logues of the substrate DHAP, or (at best) analogues of
the intermediate ene-diol(ate).
i
and 4.5 lM, respectively. Thus, 3a is as effective against
TIM as the old known phosphoglycolohydroxamate 1a.
However, one major advantage of 3a over 1a is its
increased stability towards hydrolysis: the amidoxime
remained unchanged after 72 h in water at room temper-
ature, while the hydroxamate was for part hydrolyzed
into the corresponding acid (30% as estimated by
2. Synthesis
1
5
The phosphonates 1–3b were prepared according to
Scheme 2.
1
3
C NMR). This last point is of importance in view of
a potential utilization of amidoximes as drugs, since
the release of hydroxylamine in vivo by hydrolysis
would be problematic.
The three tested inhibitors 2a, 3a, 1b give (almost)
purely competitive inhibition patterns (Figs. 2 and 3).
Of special interest the comparison between the activities
of the two inhibitors 2a and 3a on TIM and FBP-aldol-
ase (Table 1): phosphoglycolohydrazide 2a is a good
inhibitor of class II aldolases, with an only modest activ-
ity on TIM, while phosphoglycoloamidoxime 3a has
comparable performances on the two enzymes. This
means that 2a (or a prodrug derivative) could inhibit
the class II aldolase (e.g., from a pathogenic bacterium)
Scheme 2. Original synthesis of phosphonate inhibitors. Reagents: (a)
2 2 2
methyl acrylate, DBU; (b) TMS–Br; (c) NH OH; (d) NH –NH ; (e)
+
cyclohexylamine (CHA) (or: Dowex 50 CHA ); (f) acrylonitrile, DBU.
Table 1. Results of inhibition tests
Class II FBP-aldolase
TIM
IC50
K
i
IC50
K
i
a
1
2
3
1
2
a
a
a
b
b
—
—
0.01
—
—
—
—
3
0.34 ± 0.15
2.3 ± 0.19
145 ± 15
—
111 ± 38
4.5 ± 0.6
159 ± 13
—
—
—
3632 ± 54
3420 ± 73
27,000 ± 4553
9600 ± 2937
Scheme 3. New synthesis of phosphoglycoloamidoxime 3a. Reagents:
t
a) ( BuO)
3b
—
—
(
NH
2
P(O)Cl, NaH, DMF; (b) Amberlite 15, CH
3
CN; (c)
a
2
OH; (d) cyclohexylamine (CHA).
Data from Ref. 5a.

2908
M. Fonvielle et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2906–2909
Figure 4. Tautomeric forms of 1a, deprotonated 1a and 3a, compared
to the substrate DHAP and to the enolate intermediate HEI.
Figure 2. Inhibition of TIM by phosphoglycolohydrazide 2a.
either as a substrate analogue of DHAP or as a stable
analogue of the ene-diol(ate) reaction intermediate.
Since the amidoxime 3a cannot be deprotonated at
pH 7.9, and since the inhibitory properties of 1a and
7
6
5
4
3
2
1
00
00
00
00
00
00
00
0
2
0
15
3
a are quite similar (and much too low for these com-
1
0
5
0
pounds to be considered as transition-state analogues),
one can conclude that the two products are bound into
the active site of TIM in the same protonated tautomeric
form.
-
5
0
5
10 15 20 25 30 35
I] µM
[
In conclusion, we have devised the synthesis and evalu-
ation of new competitive inhibitors of rabbit muscle tri-
ose-phosphate isomerase. These compounds should also
be tested on TIM from different origins, especially from
pathogenic organisms.
I=0 µM
I=5 µM
I=10 µM
I=20 µM
I=30 µM
References and notes
1. Knowles, J. R. Nature 1991, 350, 121–124.
2. (a) Parthasarathy, S.; Eaazhisai, K.; Balaram, H.; Bala-
ram, P.; Murthy, M. R. N. J. Biol. Chem. 2003, 278,
-10
0
10
20
30
40
-
1
1
/[G3P] mM
5
2461–52470; (b) Aparicio, R.; Fereira, S. T.; Leite, N. R.;
Polikarpov, I. Acta Crystallogr., Sect. D: Biol. Crystallogr.
000, D56, 1492–1494; (c) Wieranga, R. K.; Noble, M. E.;
Figure 3. Inhibition of TIM by phosphoglycoloamidoxime 3a.
2
in vivo without affecting neither the class I aldolase nor
the TIM of the (mammal) host.
Davenport, R. C. J. Mol. Biol. 1992, 224, 1115–1126; (d)
Lambeir, A.-M.; Backman, J.; Ruiz-sanz, J.; Filimonov,
V.; Nielsen, J. E.; Kursula, I.; Norledge, B. V.; Wierenga,
R. K. Eur. J. Biochem. 2000, 267, 2516–2524; (e) Alvarez,
M.; Wouters, J.; Maes, D.; Mainfroid, V.; Rentier-Delrue,
F.; Depiereux, E.; Martial, J. A. J. Biol. Chem. 1999, 274,
From a mechanistic point of view, it is well established
that TIM proceeds via the formation of an ene-diolate
1
8
(
rather than an ene-diol). The phosphoglycolohydrox-
5
a
1
9181–19187; (f) Mande, S. C.; Mainfroid, V.; Kalk, K.
H.; Goraj, K.; Martial, J. A.; Hol, W. G. Protein Sci.
994, 3, 810–821.
amate 1a has been considered by Collins to act
through its tautomeric form as a stable analogue of an
ene-diol (Fig. 4).
1
3
4
. Hartman, F. C. Biochemistry 1971, 10, 146–154.
. Schray, K. J.; OÕConnell, E. L.; Rose, I. A. J. Biol. Chem.
1
. (a) Collins, K. D. J. Biol. Chem. 1974, 249, 136–142; (b)
Lewis, D. J.; Lowe, G. J. Chem. Soc., Chem. Commun.
1973, 713–715.
Considering the pK value of the hydroxamic acid func-
a
tion (around 9.6), other authors have estimated that 1a
can be deprotonated in the active site of TIM, and thus
could mimic the ene-diolate high-energy intermediate.
Similarly, the amidoxime 3a can be thought to be active
973, 248, 2214–2218.
5
7
6. Wolfenden, R. N. Nature 1969, 223, 704–707.

M. Fonvielle et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2906–2909
2909
ꢀ
7
8
. Lolis, E.; Petsko, G. A. Ann. Rev. Biochem. 1990, 59, 597–
30.
. Kuntz, D. A.; Osowski, R.; Schudok, M.; Wierenga, R.
K.; M u¨ ller, K.; Kessler, H.; Opperdoes, F. R. Eur. J.
Biochem. 1992, 207, 441–447.
. Velanker, S. S.; Ray, S. S.; Gokhale, R. S.; Suma, S.;
Balaram, H.; Balaram, P.; Murthy, M. R. N. Structure
(
(electrospray, M ): M + 1 = 168 g/mol.
1
Compound 3b: H NMR (D
(2H, m) (d 2.9: 2H, m; d 1.3–1.9: 10H, m; d 0.8–1.3: 10H,
6
2
0): d 1.44 (2H, m); d 2.17
13
2
m: CHA); C NMR (D 0): d 158.3, 27.5, 25.4 (50.2, 30.3,
3
25.1, 24.2, 23.7); P NMR (D
1
2
0): d 19.5, s; MS (electro-
ꢀ
9
spray, M ): M + 1 = 167 g/mol.
16. Weber, P.; Fonvielle, M.; Therisod, M. Tetrahedron Lett.
2003, 44, 9047–9049.
London) 1997, 5, 751–761.
0. Joubert, F.; Neitz, A. W.; Louw, A. I. Proteins 2001, 45,
36–143.
1. Gomez-Puyou, A.; Saavedra-Lira, E.; Becker, I.; Zubill-
aga, R. A.; Rojo-Dominguez, A.; Perez-Montfort, R.
Chem. Biol. 1995, 2, 847–855.
2. (a) Tellez-Valencia, A.; Avila-Rios, S.; Perz-Montfort, R.;
Rodriguez-Romero, A.; Tuena de Gomez-Puyou, M.;
Lopez-Calahorra, F.; Gomez-Puyou, A. Biochem. Bio-
phys. Res. Commun. 2002, 295, 958–963; (b) Willson, M.;
Callens, M.; Kuntz, D. A.; Perie, J.; Opperdoes, F. R.
Mol. Biochem. Parasitol. 1993, 59, 201–210; (c) Espinoza-
Fonseca, L. M.; Trujillo-Ferrara, J. G. Bioorg. Med.
Chem. Lett. 2004, 14, 3151–3154.
1
1
17. Synthesis of 3a: 3 mmol of anhydrous glycolonitrile
(obtained by successive azeotropic evaporations with
pyridine of an aqueous solution, Fluka) and 4.5 mmol of
di-tert-butylphosphochloridate were dissolved in 10 mL of
anhydrous DMF under argon at 0 ꢁC. NaH (9 mmol
of 60%) was added portionwise. After stirring 2 h, the
mixture was filtered, DMF was evaporated, the residue
was redissolved in dichloromethane and washed twice with
water. Di-tert-butylphosphoglycolonitrile was purified by
flash-chromatography (ethyl acetate/pentane 3:7; yield:
61%) and then deprotected by action of dry Amberlite 15
in acetonitrile at reflux for 2 h. After evaporation, the
crude phosphoglycolonitrile was treated in water with a
10 equiv excess of aqueous hydroxylamine (from Fluka).
After thorough evaporation, the residue was redissolved in
aqueous cyclohexylamine and the medium evaporated to
dryness. The white crystals of bis-cyclohexylamonium salt
3a were recrystallized in hot ethanol.
1
1
1
1
1
3. Fonvielle, M.; Weber, P.; Dabkowska, K.; Therisod, M.
Bioorg. Med. Chem. Lett. 2004, 14, 2923–2926.
4. Eubanks, L. M.; Poulter, C. D. Biochemistry 2003, 42,
1
140–1149.
5. Synthesis of 1–3b: 7.82 mmol of either methyl acrylate (!
1
, 2b) or acrylonitrile (! 3b) were added dropwise to a
Selected analytical data of new compounds:
Di-tert-butylphosphoglycolonitrile: H NMR (D
1
mixture of 5 mL of diethylphosphite and 0.78 mmol of
DBU. After 16 h, the excess of diethylphosphite was
evaporated, and the adduct was isolated through a bubble
to bubble distillation under reduced pressure (yield 70–
2
O): d 4.6
1
3
(2H, d, 10 Hz) d 1.5 (18H, s); C NMR (D O): d 114.8,
2
3
(84.4, 84.3), 50.9, 29.7; P NMR (D
1
2
0): d 10.6 (1P, t,
10 Hz).
9
0%). The phosphonate esters were cleaved by an excess of
18. (a) Davenport, R. C.; Bash, P. A.; Seaton, B. A.; Karplus,
M.; Petsko, G. A.; Ringe, D. Biochemistry 1991, 30, 5821–
5826; (b) Alagona, G.; Ghio, C.; Kollman, P. A. J. Am.
Chem. Soc. 1995, 9855–9862; (c) Aqvist, J.; Fothergill, M.
J. Biol. Chem. 1996, 271, 10010–10016.
TMS–Br. After evaporation, the residue was treated with
either a large excess of aqueous hydroxylamine (! 1b, 3b)
or 5 molar equivalents of aqueous hydrazine (! 2b). After
1
h, the reaction medium was thoroughly evaporated, and
the product finally purified as the bis-cyclohexylammo-
nium salt, recrystallized from hot ethanol.
Selected analytical data of new compounds:
19. Enzymatic assays: NADH (0.117 mM), glycerol-3-phos-
phate dehydrogenase from rabbit muscle (Fluka)
(1140 mU) and TIM from rabbit muscle (Sigma)
(20 mU) were mixed in 25 mM triethanolamine/HCl buffer
pH 7.9 (1.5 mL final volume). The cuvet was incubated
10 min at 25 ꢁC. The reaction was started by addition of
the substrate G3P (0–250 lM).
1
Compound 2b: H NMR (D 0): d 1.5 (2H, m); d 2.15 (2H,
2
m) (d 2.9: 2H, m; d 1.3–1.9: 10H, m; d 0.8–1.3: 10H, m:
1
CHA); C NMR (D
3
2
0): d 174.5, 28.3, 22.9 (50.2, 24.2,
31
3
0.2, 24.2, 23.7: CHA); P NMR (D 0): d 22.7, s; MS
2